Orange Smoothie Compressor Build Guide

[GGBB]

Here is the complete build guide for the compressor pedal I built for the 10th Anniversary contest.  Someone recently asked me for the 1590B version I had promised so I finally finished putting this all together.  Parts list, PCB transfer, and component layout are included along with lots of notes and options.  The PCB is now 1590B friendly (so technically not verified but there is very little difference between it and the original) and the guide provides instructions on how to omit any of the controls in case six knobs and a toggle is more than you want on a compressor.  Can't imagine anyone feeling that way though. 

http://www.aronnelson.com/gallery/main.php/v/diyuser/GGBB/OrangeSmoothieCompressor.pdf.html  -  Click on the "Download document" link at the top of the page.

Here's the contest entry with the original PCB layout:

http://www.diystompboxes.com/smfforum/index.php?topic=102861.msg932503#msg932503

[selfdestroyer]

Thanks, I am really looking forward to building this.

Cody

[midwayfair]

Mmm, smoooothie.

[upspoon12]

I think this compressor is just wild. Excellent build. am definitely going to try my had at this one!

[Mark Hammer]

Terrific build document, and not just because of the nod.  Carries on in the great tradition of ETI and Popular Electronics projects where everything is explained.

Thanks.

[GGBB]

I think this compressor is just wild. Excellent build. am definitely going to try my had at this one!

Cool!  Be sure to let us know how it goes.

Terrific build document, and not just because of the nod.  Carries on in the great tradition of ETI and Popular Electronics projects where everything is explained.

Thanks.

Thanks Mark - for this comment as well as for the inspiration behind the pedal.



I updated the doc a few days ago to fix a couple of typos and add a few minor build notes that had been left out.  I also changed a couple of values based on having lived with this pedal for a while now.  Nothing major - I lowered the brightness cap C9 from 33n to 10n as I was finding a little too much mid being boosted for my tastes (but some might like that).  I also lowered the band-pass cap C15 from 220n to 100n.  The 220n was pretty subtle.  I'll have to see how the 100n goes - it might be too low - I may end up settling on 150n.  As I mention in the guide, socket this cap and try different values to find what you like.  I also changed the layout ever so slightly to accomodate an additional optional cap C18 for Vbias noise filtering.  I don't have this cap or C17 (Vref noise filter) on my build and it's still very quiet, so they are both optional.

I also added a caution about doing this in a 1590B or 1590BB.  Even though the PCB is small enough, it might be a challenge fitting it stacked below the pots due to the shallowness of those enclosures.  Watch your component heights - especially the vertical trimmers, caps, and anything socketed - and measure/plan before drilling.

[upspoon12]

are you in any way interested in running a couple of the pcb's to sell at all? i am not yet set up to begin etching and making my own.

I take it its the BB Size that  you used in the one in the photo's?

[GGBB]

are you in any way interested in running a couple of the pcb's to sell at all? i am not yet set up to begin etching and making my own.

I take it its the BB Size that  you used in the one in the photo's?

The photo shows the original PCB layout which was sized for a 1590N1 enclosure.  So it is 2.4" wide whereas the 1590B sized layout in the build guide is 2.2" wide.  

<EDIT>
The 2.4" layout has a slight disadvantage because the lead spacing for the two 47n caps C1 and C4 is 0.3"/7.5mm which was done because those were the caps I happened to have on hand at the time (if someone wants a couple - I have a few left and would be happy to send them your way free of charge - PM me your details).
</EDIT>

I don't do a lot of etching myself, so I wouldn't be able to turn out PCBs very quickly just due to personal time constraints.  However, I would consider doing it if there are a few people interested and they can wait a few weeks.  But maybe somebody else who etches routinely would be willing - anyone reading this want to volunteer?

Otherwise, can you do perf board?  The original layout *could* be done on perf board by someone who's up for a challenge - I don't think it would be a terribly easy one.  Building my first two pedals on perf board is what drove me to start etching my own boards.  

[upspoon12]

haha.

i can certainly appreciate that, i try to stay away from perf boards at all costs. And i can certainly understand your time being valuable and not doing much etching yourself. I figured i'd ask anyways (never know if you had an extra or two lying around form the contest you'd wanna part with or something) I am however in absolutly no hurry. I have just yet to attempt etching let alone hone the skills well enough to consider attempting this one on my own at the moment (need the equipment and chemicals) but if others are interested i certainly would be in the mean time! The size of the pedal really doesn't bother me none, however, if you think the 1590B is more appropriate than the N1 i would certainly take your word for it.
Between my buddy and his studio and my love for pedals i've seen them come and go and i've really taken a liking to the orange smoothie. really well done, again! Cheers!

[midwayfair]

GGBB, would you (or others?) be interested in an Eagle PCB layout for this?

[GGBB]

if you think the 1590B is more appropriate than the N1 i would certainly take your word for it.

The guide has the newest layout which will fit in a 1590B.  I recommend that layout no matter what size enclosure you go with.  But I also recommend a bigger box than the 1590B, especially if you are going with all six knobs and the toggle switch.  I think the 1590B is doable, but not for the inexperienced or faint of heart like me. 

[GGBB]

GGBB, would you (or others?) be interested in an Eagle PCB layout for this?

That would be awesome Jon!  I was just thinking that rather than spending a day etching and drilling a handful of boards I would rather spend the time doing an Eagle layout for it that anyone could use to order their own PCBs.  Then I remembered that I really don't know how to use Eagle properly yet (despite having made two or three poor attempts in the past).  If you are offering, it would be greatly appreciated by me at least, and I'm sure I would order a few boards for myself (OSH Park?).

[upspoon12]

Great! good to know!

I've put together a few boxes in the past two years but I'm definitely not at the caliber where i can envision it without doing it 25 times and getting discouraged 156 times in the process. haha. And i certainly am a fan of keeping true to the full design you have here.

Certainly if midwayfair is offering i would as well be in for 2 or 3 of them as well.

Have you ever toyed with using some vintage components in your builds at all? i was considering building a couple simple boosts with old vintage capacitors (tested and within spec) on the inputs and output cap spots i wonder if this would add or take away anything to the pedal?

[GGBB]

Have you ever toyed with using some vintage components in your builds at all? i was considering building a couple simple boosts with old vintage capacitors (tested and within spec) on the inputs and output cap spots i wonder if this would add or take away anything to the pedal?

I'm not the right person to ask that - I've never tried any vintage/mojo components in any of my builds other than a few exotic op-amps.  I tend to be skeptical of that sort of thing, but until I actually try them out for myself I not in any position to say they don't do anything.  As always, the best thing to do is actually try it.

[haveyouseenhim]

I'm building this at the moment. The only thing I'm doing different is adding a blend knob like on the sampled comp.

[tubegeek]

GGBB:

I've spotted an error in the build doc. Despite that, I'd say you get 4 stars for that documentation - really thorough and certainly incites me to want to build that circuit. So thanks!

There are two places where you describe the effect of changing C9 (in the brightness boost section) and they contradict each other.

The description on p. 4 is correct:

"A larger value will lower the cut-off point allowing more of the mid-range frequencies to be boosted along with the treble."

On p.7 you have this:

"Changing the value of C9 affects the frequency range boosted. Lowering the value will introduce more mid-range frequencies into the brightness boost."
This should actually read, "Raising the value ..." or maybe even "Lowering the cutoff frequency by raising the value..."

This filter section works like this as I understand it:

Signal out of the output cap C12 enters the circuit on its way to the output jack. I'll use the HIGHLY RECOMMENDED simplification that a cap can be represented by a short circuit at high frequencies and as an open circuit at low frequencies. (Then, the actual value of the cap sets the frequency where the transition between "High" and "Low" occurs, in conjunction with any other impedances in the circuit.) I'll also examine only the extreme positions of the Brightness control.

Then we see there are only really four cases that we need to analyze to get a grip on the Brightness control's action:

1. High frequency behavior with the Brightness pot adjusted towards the C12 end
In this case C9 acts as a short circuit around the 10K Brightness pot and the result is that the output is loaded by R14=10K but not divided down by the Brightness pot/R14 combination. This is the "treble boost" behavior of the circuit which is actually just no cut.

2. Low frequency behavior with the Brightness pot adjusted towards the C12 end
In this case C9 acts as an open circuit. As a result, the circuit introduces a voltage divider between the 10K Bright pot and R14. Low frequencies are cut by a preset amount (50% voltage division.) The circuit therefore acts as a shelf cut at low frequencies, with a maximum cut of -6 dB as compared to full signal high frequencies as in Case 1.

3. High frequency behavior with the Brightness pot adjusted towards the R14 end
In this position, C9 is shorted out and has no effect. As a result, the circuit introduces a constant voltage divider between the 10K Bright pot and R14 that treats all frequencies the same.

4. Low frequency behavior with the Brightness pot adjusted towards the R14 end
Same as 3: In this position, C9 is shorted out and has no effect. As a result, the circuit introduces a constant voltage divider between the 10K Bright pot and R14 that treats all frequencies the same.

So, to sum up, all frequencies are divided down by 50% EXCEPT treble frequencies when the control is turned towards C12. The effect is to give a boost of 6 dB in the treble as compared to the lows.

The final step is to examine the effect of changing the value of C9.

If C9 is made larger, essentially the effect is to lower the frequency at which the circuit transitions from "Low Frequency" behavior to "High frequency" behavior: in other words, the frequency of the shelf transition described in Cases 1 & 2. It'll have no effect on the behavior with the control fully off, which is represented by Cases 3 & 4.

Throwing some math at the situation, and (WARNING: HANDWAVING ALERT!) ignoring any interaction with C12, we get a cutoff frequency for the original values of 10K and 10n of

f= 1/(2π*10000*.00000001) = 1591 Hz.

If we change the cap to 22n, say, the transition occurs at

f=1/(2π*10000*.000000022) = 723 Hz.

Bass cut (or, the impression of treble boost) will begin at a lower frequency, the larger we set the value of C9. Or to quote page 4, "A larger value will lower the cut-off point allowing more of the mid-range frequencies to be boosted along with the treble."

Having gone through all this, let me say that I've noticed that the output impedance of this effect is a bit high, 5K ohms for cases 2, 3 & 4, and varies with frequency as the transition to Case 1 is made by turning the control towards C12. For this reason I'd expect it to have some possibly-significant interaction differences depending on what came next in the signal chain and so I would suggest one more buffer after the Brightness circuit if that becomes an issue for any particular setup. It's kind of too bad if you'd have to do that because the current implementation with only one common-as-dirt dual opamp is really elegant as all hell.

You are absolutely welcome to include any or all of this blather in the build doc if you think it'd be helpful but I do not expect it or request it by any means, it's kind of heavy sledding and not really important, the correction to P. 7 is all i wanted to point out in case the contradiction were to confuse anybody. But since I went through it all for my own deconfusion process I thought I'd post it up.

[GGBB]

On p.7 you have this:

"Changing the value of C9 affects the frequency range boosted. Lowering the value will introduce more mid-range frequencies into the brightness boost."
This should actually read, "Raising the value ..." or maybe even "Lowering the cutoff frequency by raising the value..."

Thanks for catching that and bringing it to my attention - I'll get a corrected version up soon.

[tubegeek]

No worries. Again, thanks for doing a HELLUVA job on the documentation, truly a super project!

FYI I was looking at that C9 because I thought it might be worthwhile to add Yet One More Control: just a switch to parallel another 10nF with C9 to move the boost curve down an octave.

Or another idea might be to split it into two 4n7's and move it UP an octave.

Of course, leaving well enough alone is an option, too!

[GGBB]

I've corrected the error noted above and updated the build doc.

[bwanasonic]

GGBB, would you (or others?) be interested in an Eagle PCB layout for this?

This could also lead to a shared OSH park project, making PCBs easily available and cheap.